In OLS regression studies of changes in Treasury Hill (TB) rates on anticipated money, investigators have found negative coefficients on anticipated money, basically over the period 1977-1982, which is inconsistent with the market efficiency hypothesis.\nUsing a time-varying Bayesian regression regime, we pinpoint the precise weeks in which this negative response occurred. Furthermore, we find evidence supporting the idea that the Fed’s new monetary policy was creating more uncertainty in participant’s money supply forecasts, spurring them to adopt error-correcting procedures to modify such.\nEven though, we do not identify the precise for, of the error adaptive process, our confirmed hypothesis is consistent with the view that some kind of a process was a response to the Fed’s October, 1979 pronouncement of the new approach to monetary policy.